The Man Who Attacked His Own Numbers

In April 1968 the government of Pakistan was celebrating. It called the previous ten years the development decade, and it had the figures to justify the name. The economy had gone from growing at around three percent a year to nearly seven. Foreign economists flew in to study how it had been done. The man who had done most of it was the chief economist of the Planning Commission, Mahbub ul Haq, who had spent years travelling the country making the case for exactly these policies.

That month, in Karachi, he stood up and dismantled them.

Twenty-two families, he said, had come to control sixty-six percent of the country's industrial assets and around eighty percent of its banking and insurance. The growth was real. He was not disputing his own arithmetic. He was saying that the arithmetic had been describing something other than what everyone believed it described.

This was not a critic breaking in from outside. It was the architect, standing inside his own building, telling the room the foundations were not where the drawings said.

He never got over it. Everything he built afterwards came out of that afternoon.

The number he needed

Twenty-one years later Haq was out of government and had an idea and no institution. In 1989, on Robert McNamara's encouragement, he went to see William Draper, who ran the United Nations Development Programme. The first meeting failed. A second was arranged quickly, and that one produced Draper's backing.

The instrument that now shapes ministry priorities in a hundred countries was turned down in the room, and survived on a second appointment.

Haq then needed Amartya Sen. They had known each other since Cambridge in the mid-1950s, and had argued with each other for thirty years and enjoyed it. Sen was living in Finland. Haq telephoned him through the summer, asking him to help build a single figure that would stand against GDP.

Sen said no. You cannot compress a human life into one decimal.

What makes the exchange remarkable is that Haq agreed with him. He did not defend the index as accurate. He wanted, in his own words, a number of the same order of vulgarity as GNP, but not as blind as GNP to the social dimensions of human lives. Without it, he argued, people would admire a fuller account of development, nod respectfully, and then reach for GNP anyway, because crude was convenient.

Sen lost the argument to a man who had conceded the point.

There was one further move, and it was the shrewdest thing Haq ever did. He persuaded Draper to publish the result not as a United Nations document but as a free-standing report. Member states edit United Nations documents. They could not edit this one. He built the weapon and walled it in before firing it.

What it measured

The first Human Development Report appeared in 1990. Its opening line, borrowing Adam Smith's title in order to turn it over, was that people are the real wealth of a nation.

The Human Development Index combined three things into one figure: how long people live, how much schooling they get, and what they earn. It could stand beside GDP and disagree with it, and the first report went out of its way to show where it did.

Sri Lanka, on four hundred dollars a head, had a life expectancy of seventy-one years and an adult literacy rate of eighty-seven percent. Saudi Arabia, on six thousand two hundred, had a life expectancy of sixty-four and a literacy rate of fifty-five.

Read that column of incomes downward, then read the one beside it. Kuwait, elsewhere in the same report, had more doctors per person than Switzerland and four times the infant mortality.

The report named names. Sri Lanka, Chile, Costa Rica, Jamaica, Tanzania and Thailand ranked far higher on human development than on income. Oman, Gabon, Saudi Arabia, Algeria, Mauritania, Senegal, Cameroon and the United Arab Emirates ranked far lower. A hundred and thirty countries were listed, and the two orderings did not match.

The minister now had two numbers. He could look up twice.

What happened when he did

Draper opened the 1990 report by writing that an irresistible wave of human freedom was sweeping across many lands. The Berlin Wall had come down six months earlier. And buried in the same volume was an admission: no simple quantitative measure of human freedom had yet been designed.

So in 1991 they designed one. It was assembled quickly, from a British analyst's existing survey of forty freedoms, and scored on a blunt yes or no where the original had used four levels. The report set out the method plainly: freedom guaranteed, or freedom violated, one or zero.

The flattening broke it. Not one of the eighty-eight countries surveyed observed all forty freedoms. Sweden and Denmark, at the very top of the table, were each marked down on two. And on the criterion of political and legal equality for women, contemporary reviewers working through the table found that every single country had scored zero. On that measure, an index built to distinguish free societies from unfree ones had just reported that Sweden and Iraq were the same place.

And the data was frozen at the first of January, 1986.

Draper's foreword to the 1991 report opens on the surge of hope with which the decade began. Democracy sweeping across Eastern Europe and the Soviet Union. The Berlin Wall down. Germany reunited.

Twenty pages later, the table still scores the German Democratic Republic.

Here it is, top and bottom, out of forty points.

East Germany had ceased to exist seven months before the report went to press. The Soviet Union would be gone by Christmas. The index marked several of those countries with a small plus sign, to note that something had moved since 1985, and left the scores exactly as they were.

Eighteen countries made the high freedom band. Seventeen of them were wealthy and Western. The exception was Costa Rica.

The reaction was not academic. The agency publishing the ranking also distributed more than a billion dollars a year in grants. Malaysia's prime minister called the index racially biased. China told the agency to stay out of ideology. The Group of 77, speaking for a hundred and twenty-eight countries, demanded it be dropped.

Almost nobody attacked the scoring, which was where the actual fault lay. They attacked the politics, because that is what a number invites.

Here is the uncomfortable part. Haq's whole case to Sen was that only a single figure would make people with power pay attention. He was right. Nobody fought over the report's two hundred pages of analysis. They fought over one column of numbers, because it was the only part with force.

The force ran straight back at him.

What neither number asks

The index itself never settled. The question of how much an extra dollar should count once a country is already rich was answered and re-answered for twenty years. The treatment of income alone was rewritten repeatedly between 1990 and 2010, moving from a hard ceiling, to steepening discounts above a threshold, to a logarithm, and finally to no ceiling at all. Each answer was defensible, which is precisely the problem. None was forced. Somebody chose, and countries rose and fell in the rankings without anything changing inside them.

But the deeper gap is not in the arithmetic.

Look at what the three dimensions describe. Together they map the range of lives genuinely open to a person. Measuring that was hard won and it is not the failure.

The failure sits between the two instruments. GDP measures creation without measuring the human condition. The Human Development Index measures the human condition without measuring creation. Neither asks whether the two have anything to do with each other.

Picture two economies. In the first, people build relentlessly, new firms and new products and new ways of doing ordinary things, and are exhausted, insecure, and unable to say their own lives are going well. In the second, people are healthy, well schooled and comfortable, and almost nothing new is attempted by anyone. Both can post a respectable figure on the measures we currently keep. Neither is a place you would choose.

The failure in each case is not a missing quantity. It is a missing relationship.

Ninety years after Kuznets warned us about the limits of the number he helped create, the question is no longer whether GDP is enough. We know it is not. It is not even whether human development is enough.

We have two numbers. Neither of them was watching.

What is missing may not be a third number set beside the other two. It may be a number that measures what neither of them can. Not what a country produces, and not what it provides, but whether the people doing the building are the people whose lives are going well. Whether an economy flourishes as it creates.

Haq's number made ministers look up. The one still missing would make them stop, and ask what the growth was for.

A footnote, for now. In 1960 a young Pakistani economist named Mahbub ul Haq arrived at Harvard for a postdoctoral year. Simon Kuznets had joined the faculty months earlier. Whether the two ever met is unclear.

Thirty years later, Haq would pick up half of what Kuznets left on the table. The other half is still lying there.

Second in the series The Measure of Nations. Part One, "The Number That Came With a Warning," examined the warning Simon Kuznets attached to the number he built.